I’m sure WUWT readers will recall this excellent guest post at WUWT just over one year ago:
Now published in E&E Volume 21, Number 4 / August 2010
The thunderstorm thermostat hypothesis: How clouds and thunderstorms control the Earth’s temperature
Authors
Willis Eschenbach
Abstract
The Thunderstorm Thermostat Hypothesis is the hypothesis that tropical clouds and thunderstorms actively regulate the temperature of the earth. This keeps the earth at an equilibrium temperature regardless of changes in the forcings. Several kinds of evidence are presented to establish and elucidate the Thermostat Hypothesis-historical temperature stability of the Earth, theoretical considerations, satellite photos, and a description of the equilibrium mechanism.
See it here, PDF is available (£18.00 worthwhile to support E&E in my opinion). Or, read the WUWT version here:
I think it would be fun to boycott certain journals which have stonewalled papers which don’t fit the AGW agenda and show E&E support in the future for their open and eclectic policy.
I bought Craig Loehle’s book on becoming a successful scientist today and I’m looking forward to a good read..
Amino Acids in Meteorites says:
July 25, 2010 at 1:27 pm
R. Gates
You did not address this:
………………………………………………………………………………………………………………..
Your bitterness toward those who think the AGW hypothesis is likely correct is overpowering.
I ask that you publicly retract this lie.
______________
I find this comment you made about me to be tinged with bitterness, especially when jumped in when I was addressing Willis about a different topic, and not you:
“Instead you come up with all kinds of poor arguments to defend it. This shows everything about you…
Also, in past posts you’ve added the label “catastrophic” to my belief that the AGW model is likely correct. I resent this adjective.
Perhaps your comments toward me are only “caustic” and not bitter, in which case, I apologize for the poor choice. In general, I avoid posting to you and I would prefer you don’t post anything to me, especially when it tries to “hint” at your perception of my character…using such phrases as “shows everything about you…”
You know absolutely nothing about me, and you ought not post statements hinting that you do. If you can stick to the science and the facts, and make no comments about me personally, I will do the same for you, and that will be best for both of us.
___________
I really did want you to answer this so I could know your view of the PIOMAS graph:
are there limits to the accuracy of the PIOMAS graph?
_________
PIOMAS is a model, and as such is subject all the inaccuracies inherent in models. It will be improved upon by the upcoming CryoSat 2 data, as will everyone’s models.
H.R. said on July 24, 2010 at 7:29 pm:
If nothing else, it would be the humidity. Cold glaciated areas tend to have very low atmospheric humidity, the H2O would rather be in solid form. Offhand I would expect at some point with growing worldwide glaciation that low atmospheric humidity would greatly reduce cloud cover. Snow cover would also come into play. Snow is a diffuse cover, reflecting incoming light. With its small crystal size it would sublimate relatively quickly in an ultra-low humidity environment. Then the ice would be exposed. As is noted with glaciers where calving is occurring, light will penetrate glacial ice, and thick ice is not clear but has a bluish color therefore it absorbs some amount of visible light. With very low humidity, without snow cover, and with greatly reduced cloud cover, it seems likely sufficient sunlight would penetrate to permit a slow warming.
Eventually, after a very long warming, glaciers would stop growing, humidity will rise, more clouds will form, precipitation will occur. With rain at the edges of the glaciated areas, the rain will transfer heat and become runoff before freezing. Some will also go into any cracks at the edge, expanding and freezing thus worsening the cracks leading to calving. Net result is edge loss. This effect will be greater than the reduced insolation due to increased cloud cover as the loss will occur relatively fast. Then comes less ice cover, more exposed area soaking up more light, more warming… Interglacial occurs.
(And yes, water vapor is a potent “greenhouse gas” so as the humidity rises there will be greater retention of heat, promoting faster ice loss. But water vapor is the only “greenhouse gas” in this possible description, no CO2 increases need be involved.)
Not only an excellent hypothesis, Willis, but good of you to call it by its correct term, that is “hypothesis”, rather than the term so often wrongly used nowadays, “theory”.
I personally think that electrostatic electricity plays a large part in not only thunderstorms, but rainfall in general, and would explain why those particles of water which clouds consist of seem to defy gravity.
But then, who am I to even dare to attempt to speak of “scientific” issues? I am only a civil engineer.
R. Gates
Thank you for addressing this. I appreciate it.
My comments toward you are not caustic of bitter. They just are not softball. This is something you should expect.
If you had serious questions about the PIOMAS graph like you have serious questions about other things I would think differently of you. And I think others would too. But you find nothing wrong with it. You portray it as virtually correct with nothing other than needing small improvements. But it is clear the PIOMAS graph is seriously flawed. It is making wrong assumptions, assumptions that should have been corrected by the programmers by now. It doesn’t just have “inaccuracies inherent in models”. That graph is showing something that is laughable in the real world. And that’s not being bitter. That’s not being caustic. The PIOMAS graph diverges from reality.
Climate Models are Like Ouija Boards
R Gates,
maybe you shouldn’t take my comments so personally. I write comment with readers in mind, not with you, per se, in mind.
LarryOldtimer says:
July 25, 2010 at 3:54 pm
“[…]I personally think that electrostatic electricity plays a large part in not only thunderstorms, but rainfall in general, and would explain why those particles of water which clouds consist of seem to defy gravity.[…]”
Very good, i never thought much about that but the atmosphere has a vertical voltage gradient… Some research seems to be ongoing to find out about the electric charge of a microdroplet:
http://www.opticsinfobase.org/abstract.cfm?URI=ol-30-7-759
Chris Schoneveld says:
July 25, 2010 at 9:37 am
Have you considered plate tectonics and thus the fact that the Carboneferous coal deposits in Spitsbergen originate from a time when Spitsbergen was in a near equatorial position. Also the Paleocene coals at Spitsbergen and the oil accumulations in Alaska have nothing to do with past polar climates but everything with the paleogeographical position at the time of deposition.
The same
Redwoods grew on Greenland only 450,000 years ago. A bit too recent to invoke plate tectonics. And I thought it was all ice for Greenland. Must have been warmer for far longer than the MWP, where people grew grapes far north.
And it gets cold fast. Mammoths munching on warm-weather plants in Alaska.
kadaka (KD Knoebel) says:
July 25, 2010 at 3:15 pm
in response to
H.R. said on July 24, 2010 at 7:29 pm:
Thanks for your response. The only question is how rapidly the processes in your explanation could occur. If the paleos and geologists are correct, the earth sometimes experiences a rapid descent into glaciations. All that I’ve read so far is that it is a slower climb out of them.
Wait a sec! I think I’ve just answered my own question. The H2O precipitates out rapidly as snow (increased albedo), but given relatively constant incoming solar energy, it takes a lot of crumbling and other edge effects to reduce the albedo and release H2O vapor back into the atmosphere. You can get a meter of snow in a short time (hours) during a blizzard, but it takes a long time (days or weeks or months) to melt off. Does that sound about right?
‘Open sky’ generally won’t read (does not show any reading/voltage) using an electrometer (device capable of non-contact reading on a static electric field) … so how does this work again?
.
Cause or effect (the static charge?) i.e., it causes rain or is a result of rain?
Consider the Triboelectric effect might play a part – drops hitting each other – water surface tension, drops being torn apart – or the formation of ice particles/drops and collisions?
Now … which is cause and which is the effect?
Also please not observers report _no lightning_ except upon reaching a certain heavier-level of rainfall, and, usually at the point where the cell extends higher in altitude above the freezing level …
Quick ref: http://www.nssl.noaa.gov/primer/lightning/ltg_climatology.html
.
Is this a trick question – aren’t the physical processes that are involved what really matter?
Do you think the ‘airmass’ radiates it’s energy directly into space?
Everybody else does …
Bonus question: Which air molecules are capable of IR radiation?
.
No one has contested/argued this response? Do they all acquiesce to the answer given (actually, implied, i.e. air is a poor black body radiator and it is actually the ‘land’ masses that provide the majority of cooling via their much greater black-body IR radiating ability … and it is the transport of this ‘warmed’ air to the higher latitudes where the IR radiation of the surface cools the incoming air – via circulation from the mid-level cells and further south from the Hadley cells)?
.
Casper says:
July 25, 2010 at 2:32 pm
Actually, no. I am talking about something which is completely different from a buffer system.
Thunderstorms are an emergent phenomenon of the climate system. Emergent phenomena are ephemeral. They are created, have a lifespan, and then go out of existence. In addition, emergent phenomena are self-organizing. And in certain cases, such as thunderstorms, they can even be self-sustaining.
Emergent phenomena are often associated with threshold conditions. These are conditions which must be met before the phenomenon forms or occurs. Examples would be threshold ground temperatures, air temperatures, or relative humidity. Absent these conditions, they will not form. Above the threshold, however, they form rapidly and very prolifically.
Finally, in the case of thunderstorms, what is formed is a very efficient air conditioning solar driven heat pump. A thunderstorm can cool the surface well below the temperature threshold necessary to initiate the thunderstorm formation. It is this ability to “overshoot”, to move the system to a lower temperature than the starting temperature, that makes thunderstorms the active governors of the temperature.
And to return to your initial statement, this is why what I describe is not like a buffer system. Instead, it is an active control system which can overshoot the desired state. A buffer system can’t do that.
Many thanks.
w.
Grrrr …
Chris R. says:
July 25, 2010 at 2:47 pm
In my opinion, we are only starting to understand the huge role that atmospheric/cloud electromagnetic fields and discharges play in the climate. And to that we have to add the geo- and helio-magnetic fields, plus of course solar winds and storms, plus coronal mass ejections …
vukcevic, we seem to be in agreement:
1) The lower correlation with PDO is not surprising given that the teleconnection is atmospheric.
2) AMO & PDO are related but there is a 1/4-cycle phase-difference.
_Jim says:
July 25, 2010 at 8:55 pm
Gnomish July 24, 2010 at 10:00 pm says:
Jim….
Where do you think the heat goes when vapor condenses in a cloud? How do you think it is radiated?
Is this a trick question – aren’t the physical processes that are involved what really matter?
Do you think the ‘airmass’ radiates it’s energy directly into space?
YES, Jim. I know it for a fact. Did you think conduction or convection works in a vacuum?
Everybody else does …
Bonus question: Which air molecules are capable of IR radiation?
EVERY SINGLE ONE, Jim.
Just like every last loving bit of mass in the universe, Jim.
It’s dead, Jim.
.
Willis,
How well can we ascertain changes in global cloud quantities ?
I ask because most ideas linking clouds to albedo changes (Svensmark and others) rely on changes in quantity yet it seems to me that the task is virtually impossible due to the huge variability in cloud types and heights and the speed of changes.
I have recently begun to favour the view that global cloud quantities may not vary much at all and that the real cause of global albedo changes is movement of all the cloud bands latitudinally so that the angle of incidence of solar radiation changes.
They all seem to move either poleward or equatorward in unison.
That suggestion seems to be supported by the observation that albedo started to rise again around 2000 when the jet streams started moving back equatorward. Before that albedo had been declining as the jets had moved poleward.
Doesn’t that simple proposition resolve a number of problems ?
Thanks for your answer Willis!
[i] A buffer solution is an aqueous solution consisting of a mixture of a weak acid and its conjugate base or a weak base and its conjugate acid. [/i] It has the property that the pH of the solution changes very little when a small amount of strong acid or base is added to it. Many life forms thrive only in a relatively small pH range; an example of a buffer solution is blood. But such systems are having a limited capacity and it can be described via derivative dn/dpH.
I have an impression that the system is acting as a buffer. Because it protects the life anytime and tries to keep its state at the constant value for a input range . However overshooting can take place due to limited capacity.
I’m a physical chemist…
Willis:
I have only read the WUWT version of your fascinating paper, so forgive me if any of the following is outdated.
The mechanism you describe is so obviously real that it must be at least a major contributing factor in setting planetary temperatures. Whether the effect is quite strong enough to be the full governor you envisage is less clear. I have been trying to think of a clear-cut test, but have not yet found a suitable one (unfortunately, observing daily or seasonal albedo changes only shows the mechanism is operating, not whether it is sufficient to overcome long-term changes in the inputs or boundary conditions).
Your exegesis is I think somewhat marred by a confusion over negative feedback. Even simple linear negative feedback without lag can come arbitrarily close to eliminating variations (that’s what an op-amp does). With lag or underdamping, it can even reverse the change. A governor, such as you describe, is in fact a classic case of nonlinear negative feedback. Whereas your example of friction slowing a car isn’t actually negative feedback at all – it’s a loss or damping factor. None of this directly affects the substance of your hypothesis; however, it would be relevant to those regions in which the cloud formation/thunderstorm mechanism does not reach full strength. Even where the active governor does not fully operate, the negative feedback implicit in the process will still contribute largely to stabilising temperatures against changes in irradiance.
As has been noted, there must be limits beyond which the governor will not work. A lower limit would be when the input of solar energy is insufficient to create any thunderstorms before the evening or night. We could perhaps get a handle on this by observing the latitudes above which the mechanism does not operate. This would seem to be somewhere in the 30-40 latitude range, which might suggest that a reduction in irradiance of 10-20% would be enough to turn off the governor. I would be cautious about that, though, because clearly irradiance is not the only driver, or it would operate even at the poles (due to axial tilt, the mean irradiance at midsummer at the poles is actually slightly higher than at the equator!). An upper limit will be reacheed when so many thunderclouds are formed so early in the day that there’s no room for any more. Your albedo diagrams may give a handle on that. Another possible upper limit obtains if the thunderclouds, after their active phase, fail to dissipate before morning; we might then switch to a mode with a comparatively thin stratus cover, which substantially blocks nighttime cooling, yet permits enough sunlight to filter through to maintain surface conditions; the governor would no longer work, and any residual negative feedback would be weak.
Your hypothesis does not, it seems to me, depend upon any particular theory of global circulation, so the emphasis on the tropic-polar heat engine is perhaps misplaced. As a thought experiment, consider a planet in which only the tropics exist (slice off the ends or raise them above the atmosphere like Niven’s easter-egg planet Jinx) . Your governor will operate just fine, maintaining the tropical temperature on that world even if irradiance varies over a considerable range.
I have further thoughts, but don’t want to overload a single post.
Paul Vaughan says: July 26, 2010 at 2:54 am
vukcevic, we seem to be in agreement:
1) The lower correlation with PDO is not surprising given that the teleconnection is atmospheric.
2) AMO & PDO are related but there is a 1/4-cycle phase-difference.
Agree with the “teleconnection is atmospheric” but that may not be whole story (work in progress).
There is also a strong possibility that intensity of the huge arctic storms is influenced and directed by the Arctic’s geomagnetic flux, having direct reflection on the temperatures. Additionally precipitations’ strength affects amount of the ice build-up and also salinity of the surface currents.
http://www.vukcevic.talktalk.net/PS.htm
More details here
http://www.vukcevic.talktalk.net/NFC1.htm
Congratulations, Willis, on having your paper published.
At the moment mainstream climate science does not understand enough about how the hydrological cycle produces the observed climate oscillations, which are quasi-cyclic in nature. Your paper provides a framework for a better understanding of how Earth’s energy budget changes due to the turbulence inherent in the system. This explains why climate has not warmed as much as expected over the last 30y and why we are now dipping down into a period of cooling.
The IPCC spectre of run-away climate warming has now been exorcised!
Re: H.R. on July 25, 2010 at 7:44 pm
The answers are nice and simple, and we don’t like them. Systems tend to go to a lower energy state, shedding energy to do so. A shiny glaciated Earth is at a low energy state, going as low as possible with a planetary atmosphere.
Look at the Vostok ice core data, courtesy of JoNova on her “The 800 year lag – graphed” page. What it actually shows is that the temperature repeatedly will bottom out, followed by a rapid rise. Going by my description, the humidity bottoms out, the snow cover sublimates, then the warming of the exposed ice actually goes relatively fast. It is the overall cooling that takes a long time.
There are two features to take note of, most noticeable on the larger peaks. The rate of cooling curves downward generally, it goes along so far then the rate of cooling increases. As the planet cools and more H2O gets extracted from the air, the “greenhouse effect” provided by H2O drops off, and presumably the effect is logarithmic as with CO2, so the greatest temperature increases are found with the smallest concentration increases. Likewise, as the last dregs of water vapor get removed the rate of temperature drop will get steep.
Now examine the high peaks. A frequent feature, the temperature will peak at a high point, then comes a steep drop, then comes the long cooling. This looks like the “rapid descent into glaciation” you were mentioning, which is actually something else.
This is the overheat condition, referred to as an interglacial. The ice sheets melt, there is lots of liquid water, humidity rises, there are lots of clouds and storms. The planet’s system switches over to a very dynamic state, where it efficiently sheds lots of heat into space. More energy is absorbed than in the shiny glaciated state, and more is shed, much more. Eventually what yielded the condition passes, the system remains at the highly dynamic state shedding more energy than received for awhile, cooling the planet away from the overheat condition. Then the planet can begin icing up again and the long cooling off begins.
Thus we are currently in an extreme condition, which the planet is busy trying to end. And we don’t want it to end, thus we don’t like those answers.
To toss it out there, from the Vostok data, better seen with this graph (same source), something happened about 12,500 years ago. Temperatures have moderated and stuck to a somewhat narrow band, following a step rise. What happened? Offhand, might have been the Amazonian rainforest. As talked about here on WUWT when we were discussing biochar (see comments), that area actually shows signs of ancient land management, transforming the soil itself. As mentioned in Wikipedia’s Amazon Rainforest article:
The timing fits, and that rainforest is very dynamic, soaking up more sunlight than the savannah that otherwise would be expected to be there, and releasing water vapor and heat into the atmosphere for distribution elsewhere. Granted it seems a bit of a stretch to propose that something we mere humans did over 10,000 years ago has kept us in a climate amenable to us for so long, without even realizing what we were doing. Therefore alternative explanations are welcomed, although ones proposing “The CO2 must have done it!” will be looked at decidedly askance. 😉
Stephen Wilde says:
July 26, 2010 at 5:25 am
Satellites allow us global coverage of clouds, 24/7. Analyses of microwave brightness allow us to determine the elevation of the cloud tops. Bottoms are more problematic. However, in general we have a reasonably good idea of the number and types of clouds around the globe.
When I entertain an idea like this, my first action would be to run a “back of the envelope” calculation. Suppose at the equinox that cloud coverage is 69% on a cloud band at say 10°N, and the band moves to 15°N. How much difference would that make?
The surface insolation varies roughly as the cosine of the latitude. The change from 10°N to 15°N gives a change in insolation of ~ 7 W/m2. That is a maximum, however, and is reduced by several things.
One is that oftentimes thunderstorms are nearly as tall as they are wide (10 km or 6 miles). This means that their albedo doesn’t change much with changing solar angles.
As a result, actual measurements of the tropical surface insolation don’t change much over the latitudes of the tropics. However, it does change more in the temperate zone by latitude.
In a complex situation such as the climate, it is difficult to see what the overall effect might be. Could be large or small. However, you might be on to something.